Spectral classification

[Vasco Cossa]

Hi,

I have a catalogue of galaxies with their in-band sub-band spectral data. When I assume a power low (of the form S=v^-alpha) I get a perfect fit for some of the sources, however most of them have bad fit as they present a complex shape.

I wanted to know, if there's a way of using lmfit to take into account the complex shapes of these spectral of these galaxies. 

Same of them are GHz peak sources, some don't seem to follow any trend.

Regards, 

Vasco

[Matthew Newville]

> I wanted to know, if there's a way of using lmfit to take into account the complex shapes of these spectral of these galaxies. 

With lmfit, you would make a mathematical model for the spectral shape, identify what parameters are used for that model, make an initial estimate of those parameter values, and then "fit" that model to the data -- adjusting the values of the parameters to minimize the difference between model and data.  That is basically what all "curve fitting" does.

I have no idea what the data you have is or how to parameterize it.  Ideally, you would have some model that is meant to explain the data....

  

---

From: lmfi...@googlegroups.com <lmfi...@googlegroups.com> on behalf of Vasco Cossa <coss...@gmail.com>
Sent: Thursday, April 10, 2025 4:01 AM
To: lmfit-py <lmfi...@googlegroups.com>
Subject: Spectral classification

 

--
You received this message because you are subscribed to the Google Groups "lmfit-py" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lmfit-py+u...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/lmfit-py/221950ef-c385-4395-97aa-c5599ed08596n%40googlegroups.com.